Increasing number of assigned reads and poly(A) tail length estimation

[kethselly]

Hello,

I'm trying to identify transcripts (and the genes those transcripts arise from) that are differentially polyadenylated between a control and a mutant fly line. I've isolated RNA and used ONT direct RNA sequencing from both genotypes to detect poly(A) tail lengths.

I have used the Master of Pores worksflow to align the reads to the genome (using Minimap2). The output of this can then be run through the MOP2_tails part of the workflow, which utilizes nanopolish to call poly(A) tail lengths for each read. This worked wonderfully.

The minimap2 algorithm also outputs reads that were assigned to each gene. I've been attempting to connect the reads mapped to each gene with their corresponding poly(A) tail length, so I can look at gene-by-gene poly(A) tail length estimates.

The problem I am running into is that there are a number of reads where I get either a poly(A) tail length OR an assignment to a gene, but not both. I was wondering if there's any possible way to easily output the gene the read is aligned to in the nanopolish polya output files? Or is this perhaps already done but I just missed it somehow? for my analysis I have "contig" and that gives me the chromosome the read aligns too and then I can also see the starting point of that alignment on the chromosome, but it doesn't easily give the gene name/ID.

I'm just asking because I see a fairly dramatic difference in the lengths of poly(A) tails that have been assigned a gene and those that haven't and I'm wondering what is going on? (and perhaps this would be better posted in the nanopolish github issues, so please let me know if that would be better). Thank you!

~Seth